Although population demographics of larval lampreys in streams have been studied extensively, demographics in lake environments have not. Here, we estimated survival and rates of metamorphosis for larval sea lamprey (Petromyzon marinus) populations residing in the Great Lakes near river mouths (hereafter termed lentic areas). Tagged larvae were stocked and a Bayesian multi-state tag-recovery model was used to investigate population parameters associated with tag recovery, including survival and metamorphosis probabilities. Compared to previous studies of larvae in streams, larval growth in lentic areas was substantially slower (Brody growth coefficient = 0.00132; estimate based on the recovery of six tagged larvae), survival was slightly greater (annual survival = 63%), and the length at which 50% of the larvae would be expected to metamorphose was substantially shorter (126 mm). Stochastic simulations were used to estimate the production of parasitic stage (juvenile) sea lamprey from a hypothetical population of larvae in a lentic environment. Production of juvenile sea lamprey was substantial because, even though larval growth in these environments was slow relative to stream environments, survival was high and length at metamorphosis was less. However, estimated production of juvenile sea lamprey was less for the lentic environment than for similar simulations for river environments where larvae grew faster. In circumstances where the cost to kill a larva with lampricide was equal and control funds are limited, sea lamprey control effort may be best directed toward larvae in streams with fast-growing larvae, because stream-produced larvae will most likely contribute to juvenile sea lamprey populations.